Sensor Fusion of Gyroscope and Accelerometer for Low-Cost Attitude Determination System

This paper considers the problem of attitude determination of automated carrier given only measurements from a low-cost inertial measurement unit (IMU). During systEm modeling and design, quaternion representation is chosen to represent the attitude. In ordet to correct the drift caused by the cumulative error of the gyroscope, we combine gyroscope and accelerometer data based on gradient descent and complementary filter, which can reduce the load of MeU and improve real timing by costing less computation amounts. The experimental results indicate that the low-cost system can work in real time and output stable attitude. It has practical value of engineering.